1 package DBIx::Class::Storage::DBI;
2 # -*- mode: cperl; cperl-indent-level: 2 -*-
7 use base 'DBIx::Class::Storage';
10 use Carp::Clan qw/^DBIx::Class/;
12 use DBIx::Class::Storage::DBI::Cursor;
13 use DBIx::Class::Storage::Statistics;
17 __PACKAGE__->mk_group_accessors('simple' =>
18 qw/_connect_info _dbi_connect_info _dbh _sql_maker _sql_maker_opts _conn_pid
19 _conn_tid transaction_depth _dbh_autocommit _driver_determined savepoints/
22 # the values for these accessors are picked out (and deleted) from
23 # the attribute hashref passed to connect_info
24 my @storage_options = qw/
25 on_connect_call on_disconnect_call on_connect_do on_disconnect_do
26 disable_sth_caching unsafe auto_savepoint
28 __PACKAGE__->mk_group_accessors('simple' => @storage_options);
31 # default cursor class, overridable in connect_info attributes
32 __PACKAGE__->cursor_class('DBIx::Class::Storage::DBI::Cursor');
34 __PACKAGE__->mk_group_accessors('inherited' => qw/sql_maker_class/);
35 __PACKAGE__->sql_maker_class('DBIx::Class::SQLAHacks');
40 DBIx::Class::Storage::DBI - DBI storage handler
44 my $schema = MySchema->connect('dbi:SQLite:my.db');
46 $schema->storage->debug(1);
47 $schema->dbh_do("DROP TABLE authors");
49 $schema->resultset('Book')->search({
50 written_on => $schema->storage->datetime_parser(DateTime->now)
55 This class represents the connection to an RDBMS via L<DBI>. See
56 L<DBIx::Class::Storage> for general information. This pod only
57 documents DBI-specific methods and behaviors.
64 my $new = shift->next::method(@_);
66 $new->transaction_depth(0);
67 $new->_sql_maker_opts({});
68 $new->{savepoints} = [];
69 $new->{_in_dbh_do} = 0;
77 This method is normally called by L<DBIx::Class::Schema/connection>, which
78 encapsulates its argument list in an arrayref before passing them here.
80 The argument list may contain:
86 The same 4-element argument set one would normally pass to
87 L<DBI/connect>, optionally followed by
88 L<extra attributes|/DBIx::Class specific connection attributes>
89 recognized by DBIx::Class:
91 $connect_info_args = [ $dsn, $user, $password, \%dbi_attributes?, \%extra_attributes? ];
95 A single code reference which returns a connected
96 L<DBI database handle|DBI/connect> optionally followed by
97 L<extra attributes|/DBIx::Class specific connection attributes> recognized
100 $connect_info_args = [ sub { DBI->connect (...) }, \%extra_attributes? ];
104 A single hashref with all the attributes and the dsn/user/password
107 $connect_info_args = [{
115 $connect_info_args = [{
116 dbh_maker => sub { DBI->connect (...) },
121 This is particularly useful for L<Catalyst> based applications, allowing the
122 following config (L<Config::General> style):
127 dsn dbi:mysql:database=test
134 The C<dsn>/C<user>/C<password> combination can be substituted by the
135 C<dbh_maker> key whose value is a coderef that returns a connected
136 L<DBI database handle|DBI/connect>
140 Please note that the L<DBI> docs recommend that you always explicitly
141 set C<AutoCommit> to either I<0> or I<1>. L<DBIx::Class> further
142 recommends that it be set to I<1>, and that you perform transactions
143 via our L<DBIx::Class::Schema/txn_do> method. L<DBIx::Class> will set it
144 to I<1> if you do not do explicitly set it to zero. This is the default
145 for most DBDs. See L</DBIx::Class and AutoCommit> for details.
147 =head3 DBIx::Class specific connection attributes
149 In addition to the standard L<DBI|DBI/ATTRIBUTES_COMMON_TO_ALL_HANDLES>
150 L<connection|DBI/Database_Handle_Attributes> attributes, DBIx::Class recognizes
151 the following connection options. These options can be mixed in with your other
152 L<DBI> connection attributes, or placed in a seperate hashref
153 (C<\%extra_attributes>) as shown above.
155 Every time C<connect_info> is invoked, any previous settings for
156 these options will be cleared before setting the new ones, regardless of
157 whether any options are specified in the new C<connect_info>.
164 Specifies things to do immediately after connecting or re-connecting to
165 the database. Its value may contain:
171 This contains one SQL statement to execute.
173 =item an array reference
175 This contains SQL statements to execute in order. Each element contains
176 a string or a code reference that returns a string.
178 =item a code reference
180 This contains some code to execute. Unlike code references within an
181 array reference, its return value is ignored.
185 =item on_disconnect_do
187 Takes arguments in the same form as L</on_connect_do> and executes them
188 immediately before disconnecting from the database.
190 Note, this only runs if you explicitly call L</disconnect> on the
193 =item on_connect_call
195 A more generalized form of L</on_connect_do> that calls the specified
196 C<connect_call_METHOD> methods in your storage driver.
198 on_connect_do => 'select 1'
202 on_connect_call => [ [ do_sql => 'select 1' ] ]
204 Its values may contain:
210 Will call the C<connect_call_METHOD> method.
212 =item a code reference
214 Will execute C<< $code->($storage) >>
216 =item an array reference
218 Each value can be a method name or code reference.
220 =item an array of arrays
222 For each array, the first item is taken to be the C<connect_call_> method name
223 or code reference, and the rest are parameters to it.
227 Some predefined storage methods you may use:
233 Executes a SQL string or a code reference that returns a SQL string. This is
234 what L</on_connect_do> and L</on_disconnect_do> use.
242 Will execute the scalar as SQL.
246 Taken to be arguments to L<DBI/do>, the SQL string optionally followed by the
247 attributes hashref and bind values.
249 =item a code reference
251 Will execute C<< $code->($storage) >> and execute the return array refs as
258 Execute any statements necessary to initialize the database session to return
259 and accept datetime/timestamp values used with
260 L<DBIx::Class::InflateColumn::DateTime>.
262 Only necessary for some databases, see your specific storage driver for
263 implementation details.
267 =item on_disconnect_call
269 Takes arguments in the same form as L</on_connect_call> and executes them
270 immediately before disconnecting from the database.
272 Calls the C<disconnect_call_METHOD> methods as opposed to the
273 C<connect_call_METHOD> methods called by L</on_connect_call>.
275 Note, this only runs if you explicitly call L</disconnect> on the
278 =item disable_sth_caching
280 If set to a true value, this option will disable the caching of
281 statement handles via L<DBI/prepare_cached>.
285 Sets the limit dialect. This is useful for JDBC-bridge among others
286 where the remote SQL-dialect cannot be determined by the name of the
287 driver alone. See also L<SQL::Abstract::Limit>.
291 Specifies what characters to use to quote table and column names. If
292 you use this you will want to specify L</name_sep> as well.
294 C<quote_char> expects either a single character, in which case is it
295 is placed on either side of the table/column name, or an arrayref of length
296 2 in which case the table/column name is placed between the elements.
298 For example under MySQL you should use C<< quote_char => '`' >>, and for
299 SQL Server you should use C<< quote_char => [qw/[ ]/] >>.
303 This only needs to be used in conjunction with C<quote_char>, and is used to
304 specify the charecter that seperates elements (schemas, tables, columns) from
305 each other. In most cases this is simply a C<.>.
307 The consequences of not supplying this value is that L<SQL::Abstract>
308 will assume DBIx::Class' uses of aliases to be complete column
309 names. The output will look like I<"me.name"> when it should actually
314 This Storage driver normally installs its own C<HandleError>, sets
315 C<RaiseError> and C<ShowErrorStatement> on, and sets C<PrintError> off on
316 all database handles, including those supplied by a coderef. It does this
317 so that it can have consistent and useful error behavior.
319 If you set this option to a true value, Storage will not do its usual
320 modifications to the database handle's attributes, and instead relies on
321 the settings in your connect_info DBI options (or the values you set in
322 your connection coderef, in the case that you are connecting via coderef).
324 Note that your custom settings can cause Storage to malfunction,
325 especially if you set a C<HandleError> handler that suppresses exceptions
326 and/or disable C<RaiseError>.
330 If this option is true, L<DBIx::Class> will use savepoints when nesting
331 transactions, making it possible to recover from failure in the inner
332 transaction without having to abort all outer transactions.
336 Use this argument to supply a cursor class other than the default
337 L<DBIx::Class::Storage::DBI::Cursor>.
341 Some real-life examples of arguments to L</connect_info> and
342 L<DBIx::Class::Schema/connect>
344 # Simple SQLite connection
345 ->connect_info([ 'dbi:SQLite:./foo.db' ]);
348 ->connect_info([ sub { DBI->connect(...) } ]);
350 # Connect via subref in hashref
352 dbh_maker => sub { DBI->connect(...) },
353 on_connect_do => 'alter session ...',
356 # A bit more complicated
363 { quote_char => q{"}, name_sep => q{.} },
367 # Equivalent to the previous example
373 { AutoCommit => 1, quote_char => q{"}, name_sep => q{.} },
377 # Same, but with hashref as argument
378 # See parse_connect_info for explanation
381 dsn => 'dbi:Pg:dbname=foo',
383 password => 'my_pg_password',
390 # Subref + DBIx::Class-specific connection options
393 sub { DBI->connect(...) },
397 on_connect_do => ['SET search_path TO myschema,otherschema,public'],
398 disable_sth_caching => 1,
408 my ($self, $info_arg) = @_;
410 return $self->_connect_info if !$info_arg;
412 my @args = @$info_arg; # take a shallow copy for further mutilation
413 $self->_connect_info([@args]); # copy for _connect_info
416 # combine/pre-parse arguments depending on invocation style
419 if (ref $args[0] eq 'CODE') { # coderef with optional \%extra_attributes
420 %attrs = %{ $args[1] || {} };
423 elsif (ref $args[0] eq 'HASH') { # single hashref (i.e. Catalyst config)
424 %attrs = %{$args[0]};
426 if (my $code = delete $attrs{dbh_maker}) {
429 my @ignored = grep { delete $attrs{$_} } (qw/dsn user password/);
432 'Attribute(s) %s in connect_info were ignored, as they can not be applied '
433 . "to the result of 'dbh_maker'",
435 join (', ', map { "'$_'" } (@ignored) ),
440 @args = delete @attrs{qw/dsn user password/};
443 else { # otherwise assume dsn/user/password + \%attrs + \%extra_attrs
445 % { $args[3] || {} },
446 % { $args[4] || {} },
448 @args = @args[0,1,2];
451 # Kill sql_maker/_sql_maker_opts, so we get a fresh one with only
452 # the new set of options
453 $self->_sql_maker(undef);
454 $self->_sql_maker_opts({});
457 for my $storage_opt (@storage_options, 'cursor_class') { # @storage_options is declared at the top of the module
458 if(my $value = delete $attrs{$storage_opt}) {
459 $self->$storage_opt($value);
462 for my $sql_maker_opt (qw/limit_dialect quote_char name_sep/) {
463 if(my $opt_val = delete $attrs{$sql_maker_opt}) {
464 $self->_sql_maker_opts->{$sql_maker_opt} = $opt_val;
469 if (ref $args[0] eq 'CODE') {
470 # _connect() never looks past $args[0] in this case
474 %{ $self->_default_dbi_connect_attributes || {} },
479 $self->_dbi_connect_info([@args, keys %attrs ? \%attrs : ()]);
480 $self->_connect_info;
483 sub _default_dbi_connect_attributes {
493 This method is deprecated in favour of setting via L</connect_info>.
497 =head2 on_disconnect_do
499 This method is deprecated in favour of setting via L</connect_info>.
503 sub _parse_connect_do {
504 my ($self, $type) = @_;
506 my $val = $self->$type;
507 return () if not defined $val;
512 push @res, [ 'do_sql', $val ];
513 } elsif (ref($val) eq 'CODE') {
515 } elsif (ref($val) eq 'ARRAY') {
516 push @res, map { [ 'do_sql', $_ ] } @$val;
518 $self->throw_exception("Invalid type for $type: ".ref($val));
526 Arguments: ($subref | $method_name), @extra_coderef_args?
528 Execute the given $subref or $method_name using the new exception-based
529 connection management.
531 The first two arguments will be the storage object that C<dbh_do> was called
532 on and a database handle to use. Any additional arguments will be passed
533 verbatim to the called subref as arguments 2 and onwards.
535 Using this (instead of $self->_dbh or $self->dbh) ensures correct
536 exception handling and reconnection (or failover in future subclasses).
538 Your subref should have no side-effects outside of the database, as
539 there is the potential for your subref to be partially double-executed
540 if the database connection was stale/dysfunctional.
544 my @stuff = $schema->storage->dbh_do(
546 my ($storage, $dbh, @cols) = @_;
547 my $cols = join(q{, }, @cols);
548 $dbh->selectrow_array("SELECT $cols FROM foo");
559 my $dbh = $self->_dbh;
561 return $self->$code($dbh, @_) if $self->{_in_dbh_do}
562 || $self->{transaction_depth};
564 local $self->{_in_dbh_do} = 1;
567 my $want_array = wantarray;
570 $self->_verify_pid if $dbh;
572 $self->_populate_dbh;
577 @result = $self->$code($dbh, @_);
579 elsif(defined $want_array) {
580 $result[0] = $self->$code($dbh, @_);
583 $self->$code($dbh, @_);
587 # ->connected might unset $@ - copy
589 if(!$exception) { return $want_array ? @result : $result[0] }
591 $self->throw_exception($exception) if $self->connected;
593 # We were not connected - reconnect and retry, but let any
594 # exception fall right through this time
595 carp "Retrying $code after catching disconnected exception: $exception"
596 if $ENV{DBIC_DBIRETRY_DEBUG};
597 $self->_populate_dbh;
598 $self->$code($self->_dbh, @_);
601 # This is basically a blend of dbh_do above and DBIx::Class::Storage::txn_do.
602 # It also informs dbh_do to bypass itself while under the direction of txn_do,
603 # via $self->{_in_dbh_do} (this saves some redundant eval and errorcheck, etc)
608 ref $coderef eq 'CODE' or $self->throw_exception
609 ('$coderef must be a CODE reference');
611 return $coderef->(@_) if $self->{transaction_depth} && ! $self->auto_savepoint;
613 local $self->{_in_dbh_do} = 1;
616 my $want_array = wantarray;
621 $self->_verify_pid if $self->_dbh;
622 $self->_populate_dbh if !$self->_dbh;
626 @result = $coderef->(@_);
628 elsif(defined $want_array) {
629 $result[0] = $coderef->(@_);
637 # ->connected might unset $@ - copy
639 if(!$exception) { return $want_array ? @result : $result[0] }
641 if($tried++ || $self->connected) {
642 eval { $self->txn_rollback };
643 my $rollback_exception = $@;
644 if($rollback_exception) {
645 my $exception_class = "DBIx::Class::Storage::NESTED_ROLLBACK_EXCEPTION";
646 $self->throw_exception($exception) # propagate nested rollback
647 if $rollback_exception =~ /$exception_class/;
649 $self->throw_exception(
650 "Transaction aborted: ${exception}. "
651 . "Rollback failed: ${rollback_exception}"
654 $self->throw_exception($exception)
657 # We were not connected, and was first try - reconnect and retry
659 carp "Retrying $coderef after catching disconnected exception: $exception"
660 if $ENV{DBIC_DBIRETRY_DEBUG};
661 $self->_populate_dbh;
667 Our C<disconnect> method also performs a rollback first if the
668 database is not in C<AutoCommit> mode.
678 push @actions, ( $self->on_disconnect_call || () );
679 push @actions, $self->_parse_connect_do ('on_disconnect_do');
681 $self->_do_connection_actions(disconnect_call_ => $_) for @actions;
683 $self->_dbh_rollback unless $self->_dbh_autocommit;
685 $self->_dbh->disconnect;
691 =head2 with_deferred_fk_checks
695 =item Arguments: C<$coderef>
697 =item Return Value: The return value of $coderef
701 Storage specific method to run the code ref with FK checks deferred or
702 in MySQL's case disabled entirely.
706 # Storage subclasses should override this
707 sub with_deferred_fk_checks {
708 my ($self, $sub) = @_;
717 =item Arguments: none
719 =item Return Value: 1|0
723 Verifies that the the current database handle is active and ready to execute
724 an SQL statement (i.e. the connection did not get stale, server is still
725 answering, etc.) This method is used internally by L</dbh>.
731 return 0 unless $self->_seems_connected;
734 local $self->_dbh->{RaiseError} = 1;
739 sub _seems_connected {
742 my $dbh = $self->_dbh
745 if(defined $self->_conn_tid && $self->_conn_tid != threads->tid) {
752 return 0 if !$self->_dbh;
755 return $dbh->FETCH('Active');
761 my $dbh = $self->_dbh or return 0;
766 # handle pid changes correctly
767 # NOTE: assumes $self->_dbh is a valid $dbh
771 return if defined $self->_conn_pid && $self->_conn_pid == $$;
773 $self->_dbh->{InactiveDestroy} = 1;
780 sub ensure_connected {
783 unless ($self->connected) {
784 $self->_populate_dbh;
790 Returns a C<$dbh> - a data base handle of class L<DBI>. The returned handle
791 is guaranteed to be healthy by implicitly calling L</connected>, and if
792 necessary performing a reconnection before returning. Keep in mind that this
793 is very B<expensive> on some database engines. Consider using L<dbh_do>
801 if (not $self->_dbh) {
802 $self->_populate_dbh;
804 $self->ensure_connected;
809 # this is the internal "get dbh or connect (don't check)" method
812 $self->_populate_dbh unless $self->_dbh;
816 sub _sql_maker_args {
821 array_datatypes => 1,
822 limit_dialect => $self->_get_dbh,
823 %{$self->_sql_maker_opts}
829 unless ($self->_sql_maker) {
830 my $sql_maker_class = $self->sql_maker_class;
831 $self->ensure_class_loaded ($sql_maker_class);
832 $self->_sql_maker($sql_maker_class->new( $self->_sql_maker_args ));
834 return $self->_sql_maker;
842 my @info = @{$self->_dbi_connect_info || []};
843 $self->_dbh(undef); # in case ->connected failed we might get sent here
844 $self->_dbh($self->_connect(@info));
846 $self->_conn_pid($$);
847 $self->_conn_tid(threads->tid) if $INC{'threads.pm'};
849 $self->_determine_driver;
851 # Always set the transaction depth on connect, since
852 # there is no transaction in progress by definition
853 $self->{transaction_depth} = $self->_dbh_autocommit ? 0 : 1;
855 $self->_run_connection_actions unless $self->{_in_determine_driver};
858 sub _run_connection_actions {
862 push @actions, ( $self->on_connect_call || () );
863 push @actions, $self->_parse_connect_do ('on_connect_do');
865 $self->_do_connection_actions(connect_call_ => $_) for @actions;
868 sub _determine_driver {
871 if ((not $self->_driver_determined) && (not $self->{_in_determine_driver})) {
872 my $started_unconnected = 0;
873 local $self->{_in_determine_driver} = 1;
875 if (ref($self) eq __PACKAGE__) {
877 if ($self->_dbh) { # we are connected
878 $driver = $self->_dbh->{Driver}{Name};
880 # try to use dsn to not require being connected, the driver may still
881 # force a connection in _rebless to determine version
882 ($driver) = $self->_dbi_connect_info->[0] =~ /dbi:([^:]+):/i;
883 $started_unconnected = 1;
886 my $storage_class = "DBIx::Class::Storage::DBI::${driver}";
887 if ($self->load_optional_class($storage_class)) {
888 mro::set_mro($storage_class, 'c3');
889 bless $self, $storage_class;
894 $self->_driver_determined(1);
896 $self->_run_connection_actions
897 if $started_unconnected && defined $self->_dbh;
901 sub _do_connection_actions {
903 my $method_prefix = shift;
906 if (not ref($call)) {
907 my $method = $method_prefix . $call;
909 } elsif (ref($call) eq 'CODE') {
911 } elsif (ref($call) eq 'ARRAY') {
912 if (ref($call->[0]) ne 'ARRAY') {
913 $self->_do_connection_actions($method_prefix, $_) for @$call;
915 $self->_do_connection_actions($method_prefix, @$_) for @$call;
918 $self->throw_exception (sprintf ("Don't know how to process conection actions of type '%s'", ref($call)) );
924 sub connect_call_do_sql {
926 $self->_do_query(@_);
929 sub disconnect_call_do_sql {
931 $self->_do_query(@_);
934 # override in db-specific backend when necessary
935 sub connect_call_datetime_setup { 1 }
938 my ($self, $action) = @_;
940 if (ref $action eq 'CODE') {
941 $action = $action->($self);
942 $self->_do_query($_) foreach @$action;
945 # Most debuggers expect ($sql, @bind), so we need to exclude
946 # the attribute hash which is the second argument to $dbh->do
947 # furthermore the bind values are usually to be presented
948 # as named arrayref pairs, so wrap those here too
949 my @do_args = (ref $action eq 'ARRAY') ? (@$action) : ($action);
950 my $sql = shift @do_args;
951 my $attrs = shift @do_args;
952 my @bind = map { [ undef, $_ ] } @do_args;
954 $self->_query_start($sql, @bind);
955 $self->_dbh->do($sql, $attrs, @do_args);
956 $self->_query_end($sql, @bind);
963 my ($self, @info) = @_;
965 $self->throw_exception("You failed to provide any connection info")
968 my ($old_connect_via, $dbh);
970 if ($INC{'Apache/DBI.pm'} && $ENV{MOD_PERL}) {
971 $old_connect_via = $DBI::connect_via;
972 $DBI::connect_via = 'connect';
976 if(ref $info[0] eq 'CODE') {
980 $dbh = DBI->connect(@info);
983 if($dbh && !$self->unsafe) {
984 my $weak_self = $self;
985 Scalar::Util::weaken($weak_self);
986 $dbh->{HandleError} = sub {
988 $weak_self->throw_exception("DBI Exception: $_[0]");
991 croak ("DBI Exception: $_[0]");
994 $dbh->{ShowErrorStatement} = 1;
995 $dbh->{RaiseError} = 1;
996 $dbh->{PrintError} = 0;
1000 $DBI::connect_via = $old_connect_via if $old_connect_via;
1002 $self->throw_exception("DBI Connection failed: " . ($@||$DBI::errstr))
1005 $self->_dbh_autocommit($dbh->{AutoCommit});
1011 my ($self, $name) = @_;
1013 $name = $self->_svp_generate_name
1014 unless defined $name;
1016 $self->throw_exception ("You can't use savepoints outside a transaction")
1017 if $self->{transaction_depth} == 0;
1019 $self->throw_exception ("Your Storage implementation doesn't support savepoints")
1020 unless $self->can('_svp_begin');
1022 push @{ $self->{savepoints} }, $name;
1024 $self->debugobj->svp_begin($name) if $self->debug;
1026 return $self->_svp_begin($name);
1030 my ($self, $name) = @_;
1032 $self->throw_exception ("You can't use savepoints outside a transaction")
1033 if $self->{transaction_depth} == 0;
1035 $self->throw_exception ("Your Storage implementation doesn't support savepoints")
1036 unless $self->can('_svp_release');
1038 if (defined $name) {
1039 $self->throw_exception ("Savepoint '$name' does not exist")
1040 unless grep { $_ eq $name } @{ $self->{savepoints} };
1042 # Dig through the stack until we find the one we are releasing. This keeps
1043 # the stack up to date.
1046 do { $svp = pop @{ $self->{savepoints} } } while $svp ne $name;
1048 $name = pop @{ $self->{savepoints} };
1051 $self->debugobj->svp_release($name) if $self->debug;
1053 return $self->_svp_release($name);
1057 my ($self, $name) = @_;
1059 $self->throw_exception ("You can't use savepoints outside a transaction")
1060 if $self->{transaction_depth} == 0;
1062 $self->throw_exception ("Your Storage implementation doesn't support savepoints")
1063 unless $self->can('_svp_rollback');
1065 if (defined $name) {
1066 # If they passed us a name, verify that it exists in the stack
1067 unless(grep({ $_ eq $name } @{ $self->{savepoints} })) {
1068 $self->throw_exception("Savepoint '$name' does not exist!");
1071 # Dig through the stack until we find the one we are releasing. This keeps
1072 # the stack up to date.
1073 while(my $s = pop(@{ $self->{savepoints} })) {
1074 last if($s eq $name);
1076 # Add the savepoint back to the stack, as a rollback doesn't remove the
1077 # named savepoint, only everything after it.
1078 push(@{ $self->{savepoints} }, $name);
1080 # We'll assume they want to rollback to the last savepoint
1081 $name = $self->{savepoints}->[-1];
1084 $self->debugobj->svp_rollback($name) if $self->debug;
1086 return $self->_svp_rollback($name);
1089 sub _svp_generate_name {
1092 return 'savepoint_'.scalar(@{ $self->{'savepoints'} });
1097 if($self->{transaction_depth} == 0) {
1098 $self->debugobj->txn_begin()
1100 $self->_dbh_begin_work;
1102 elsif ($self->auto_savepoint) {
1105 $self->{transaction_depth}++;
1108 sub _dbh_begin_work {
1111 # if the user is utilizing txn_do - good for him, otherwise we need to
1112 # ensure that the $dbh is healthy on BEGIN.
1113 # We do this via ->dbh_do instead of ->dbh, so that the ->dbh "ping"
1114 # will be replaced by a failure of begin_work itself (which will be
1115 # then retried on reconnect)
1116 if ($self->{_in_dbh_do}) {
1117 $self->_dbh->begin_work;
1119 $self->dbh_do(sub { $_[1]->begin_work });
1125 if ($self->{transaction_depth} == 1) {
1126 my $dbh = $self->_dbh;
1127 $self->debugobj->txn_commit()
1130 $self->{transaction_depth} = 0
1131 if $self->_dbh_autocommit;
1133 elsif($self->{transaction_depth} > 1) {
1134 $self->{transaction_depth}--;
1136 if $self->auto_savepoint;
1142 $self->_dbh->commit;
1147 my $dbh = $self->_dbh;
1149 if ($self->{transaction_depth} == 1) {
1150 $self->debugobj->txn_rollback()
1152 $self->{transaction_depth} = 0
1153 if $self->_dbh_autocommit;
1154 $self->_dbh_rollback;
1156 elsif($self->{transaction_depth} > 1) {
1157 $self->{transaction_depth}--;
1158 if ($self->auto_savepoint) {
1159 $self->svp_rollback;
1164 die DBIx::Class::Storage::NESTED_ROLLBACK_EXCEPTION->new;
1169 my $exception_class = "DBIx::Class::Storage::NESTED_ROLLBACK_EXCEPTION";
1170 $error =~ /$exception_class/ and $self->throw_exception($error);
1171 # ensure that a failed rollback resets the transaction depth
1172 $self->{transaction_depth} = $self->_dbh_autocommit ? 0 : 1;
1173 $self->throw_exception($error);
1179 $self->_dbh->rollback;
1182 # This used to be the top-half of _execute. It was split out to make it
1183 # easier to override in NoBindVars without duping the rest. It takes up
1184 # all of _execute's args, and emits $sql, @bind.
1185 sub _prep_for_execute {
1186 my ($self, $op, $extra_bind, $ident, $args) = @_;
1188 if( Scalar::Util::blessed($ident) && $ident->isa("DBIx::Class::ResultSource") ) {
1189 $ident = $ident->from();
1192 my ($sql, @bind) = $self->sql_maker->$op($ident, @$args);
1195 map { ref $_ eq 'ARRAY' ? $_ : [ '!!dummy', $_ ] } @$extra_bind)
1197 return ($sql, \@bind);
1201 sub _fix_bind_params {
1202 my ($self, @bind) = @_;
1204 ### Turn @bind from something like this:
1205 ### ( [ "artist", 1 ], [ "cdid", 1, 3 ] )
1207 ### ( "'1'", "'1'", "'3'" )
1210 if ( defined( $_ && $_->[1] ) ) {
1211 map { qq{'$_'}; } @{$_}[ 1 .. $#$_ ];
1218 my ( $self, $sql, @bind ) = @_;
1220 if ( $self->debug ) {
1221 @bind = $self->_fix_bind_params(@bind);
1223 $self->debugobj->query_start( $sql, @bind );
1228 my ( $self, $sql, @bind ) = @_;
1230 if ( $self->debug ) {
1231 @bind = $self->_fix_bind_params(@bind);
1232 $self->debugobj->query_end( $sql, @bind );
1237 my ($self, $dbh, $op, $extra_bind, $ident, $bind_attributes, @args) = @_;
1239 my ($sql, $bind) = $self->_prep_for_execute($op, $extra_bind, $ident, \@args);
1241 $self->_query_start( $sql, @$bind );
1243 my $sth = $self->sth($sql,$op);
1245 my $placeholder_index = 1;
1247 foreach my $bound (@$bind) {
1248 my $attributes = {};
1249 my($column_name, @data) = @$bound;
1251 if ($bind_attributes) {
1252 $attributes = $bind_attributes->{$column_name}
1253 if defined $bind_attributes->{$column_name};
1256 foreach my $data (@data) {
1257 my $ref = ref $data;
1258 $data = $ref && $ref ne 'ARRAY' ? ''.$data : $data; # stringify args (except arrayrefs)
1260 $sth->bind_param($placeholder_index, $data, $attributes);
1261 $placeholder_index++;
1265 # Can this fail without throwing an exception anyways???
1266 my $rv = $sth->execute();
1267 $self->throw_exception($sth->errstr) if !$rv;
1269 $self->_query_end( $sql, @$bind );
1271 return (wantarray ? ($rv, $sth, @$bind) : $rv);
1276 $self->dbh_do('_dbh_execute', @_); # retry over disconnects
1280 my ($self, $source, $to_insert) = @_;
1282 # redispatch to insert method of storage we reblessed into, if necessary
1283 if (not $self->_driver_determined) {
1284 $self->_determine_driver;
1285 goto $self->can('insert');
1288 my $ident = $source->from;
1289 my $bind_attributes = $self->source_bind_attributes($source);
1291 my $updated_cols = {};
1293 foreach my $col ( $source->columns ) {
1294 if ( !defined $to_insert->{$col} ) {
1295 my $col_info = $source->column_info($col);
1297 if ( $col_info->{auto_nextval} ) {
1298 $updated_cols->{$col} = $to_insert->{$col} = $self->_sequence_fetch(
1300 $col_info->{sequence} ||
1301 $self->_dbh_get_autoinc_seq($self->_get_dbh, $source)
1307 $self->_execute('insert' => [], $source, $bind_attributes, $to_insert);
1309 return $updated_cols;
1312 ## Still not quite perfect, and EXPERIMENTAL
1313 ## Currently it is assumed that all values passed will be "normal", i.e. not
1314 ## scalar refs, or at least, all the same type as the first set, the statement is
1315 ## only prepped once.
1317 my ($self, $source, $cols, $data) = @_;
1319 my $table = $source->from;
1320 @colvalues{@$cols} = (0..$#$cols);
1321 my ($sql, @bind) = $self->sql_maker->insert($table, \%colvalues);
1323 $self->_determine_driver;
1325 $self->_query_start( $sql, @bind );
1326 my $sth = $self->sth($sql);
1328 # @bind = map { ref $_ ? ''.$_ : $_ } @bind; # stringify args
1330 ## This must be an arrayref, else nothing works!
1331 my $tuple_status = [];
1333 ## Get the bind_attributes, if any exist
1334 my $bind_attributes = $self->source_bind_attributes($source);
1336 ## Bind the values and execute
1337 my $placeholder_index = 1;
1339 foreach my $bound (@bind) {
1341 my $attributes = {};
1342 my ($column_name, $data_index) = @$bound;
1344 if( $bind_attributes ) {
1345 $attributes = $bind_attributes->{$column_name}
1346 if defined $bind_attributes->{$column_name};
1349 my @data = map { $_->[$data_index] } @$data;
1351 $sth->bind_param_array( $placeholder_index, [@data], $attributes );
1352 $placeholder_index++;
1354 my $rv = eval { $sth->execute_array({ArrayTupleStatus => $tuple_status}) };
1357 ++$i while $i <= $#$tuple_status && !ref $tuple_status->[$i];
1359 $self->throw_exception($sth->errstr || "Unexpected populate error: $err")
1360 if ($i > $#$tuple_status);
1362 require Data::Dumper;
1363 local $Data::Dumper::Terse = 1;
1364 local $Data::Dumper::Indent = 1;
1365 local $Data::Dumper::Useqq = 1;
1366 local $Data::Dumper::Quotekeys = 0;
1368 $self->throw_exception(sprintf "%s for populate slice:\n%s",
1369 $tuple_status->[$i][1],
1370 Data::Dumper::Dumper(
1371 { map { $cols->[$_] => $data->[$i][$_] } (0 .. $#$cols) }
1375 $self->throw_exception($sth->errstr) if !$rv;
1377 $self->_query_end( $sql, @bind );
1378 return (wantarray ? ($rv, $sth, @bind) : $rv);
1382 my ($self, $source, @args) = @_;
1384 # redispatch to update method of storage we reblessed into, if necessary
1385 if (not $self->_driver_determined) {
1386 $self->_determine_driver;
1387 goto $self->can('update');
1390 my $bind_attributes = $self->source_bind_attributes($source);
1392 return $self->_execute('update' => [], $source, $bind_attributes, @args);
1397 my $self = shift @_;
1398 my $source = shift @_;
1399 $self->_determine_driver;
1400 my $bind_attrs = $self->source_bind_attributes($source);
1402 return $self->_execute('delete' => [], $source, $bind_attrs, @_);
1405 # We were sent here because the $rs contains a complex search
1406 # which will require a subquery to select the correct rows
1407 # (i.e. joined or limited resultsets)
1409 # Genarating a single PK column subquery is trivial and supported
1410 # by all RDBMS. However if we have a multicolumn PK, things get ugly.
1411 # Look at _multipk_update_delete()
1412 sub _subq_update_delete {
1414 my ($rs, $op, $values) = @_;
1416 my $rsrc = $rs->result_source;
1418 # we already check this, but double check naively just in case. Should be removed soon
1419 my $sel = $rs->_resolved_attrs->{select};
1420 $sel = [ $sel ] unless ref $sel eq 'ARRAY';
1421 my @pcols = $rsrc->primary_columns;
1422 if (@$sel != @pcols) {
1423 $self->throw_exception (
1424 'Subquery update/delete can not be called on resultsets selecting a'
1425 .' number of columns different than the number of primary keys'
1432 $op eq 'update' ? $values : (),
1433 { $pcols[0] => { -in => $rs->as_query } },
1438 return $self->_multipk_update_delete (@_);
1442 # ANSI SQL does not provide a reliable way to perform a multicol-PK
1443 # resultset update/delete involving subqueries. So by default resort
1444 # to simple (and inefficient) delete_all style per-row opearations,
1445 # while allowing specific storages to override this with a faster
1448 sub _multipk_update_delete {
1449 return shift->_per_row_update_delete (@_);
1452 # This is the default loop used to delete/update rows for multi PK
1453 # resultsets, and used by mysql exclusively (because it can't do anything
1456 # We do not use $row->$op style queries, because resultset update/delete
1457 # is not expected to cascade (this is what delete_all/update_all is for).
1459 # There should be no race conditions as the entire operation is rolled
1462 sub _per_row_update_delete {
1464 my ($rs, $op, $values) = @_;
1466 my $rsrc = $rs->result_source;
1467 my @pcols = $rsrc->primary_columns;
1469 my $guard = $self->txn_scope_guard;
1471 # emulate the return value of $sth->execute for non-selects
1472 my $row_cnt = '0E0';
1474 my $subrs_cur = $rs->cursor;
1475 while (my @pks = $subrs_cur->next) {
1478 for my $i (0.. $#pcols) {
1479 $cond->{$pcols[$i]} = $pks[$i];
1484 $op eq 'update' ? $values : (),
1499 # localization is neccessary as
1500 # 1) there is no infrastructure to pass this around before SQLA2
1501 # 2) _select_args sets it and _prep_for_execute consumes it
1502 my $sql_maker = $self->sql_maker;
1503 local $sql_maker->{_dbic_rs_attrs};
1505 return $self->_execute($self->_select_args(@_));
1508 sub _select_args_to_query {
1511 # localization is neccessary as
1512 # 1) there is no infrastructure to pass this around before SQLA2
1513 # 2) _select_args sets it and _prep_for_execute consumes it
1514 my $sql_maker = $self->sql_maker;
1515 local $sql_maker->{_dbic_rs_attrs};
1517 # my ($op, $bind, $ident, $bind_attrs, $select, $cond, $order, $rows, $offset)
1518 # = $self->_select_args($ident, $select, $cond, $attrs);
1519 my ($op, $bind, $ident, $bind_attrs, @args) =
1520 $self->_select_args(@_);
1522 # my ($sql, $prepared_bind) = $self->_prep_for_execute($op, $bind, $ident, [ $select, $cond, $order, $rows, $offset ]);
1523 my ($sql, $prepared_bind) = $self->_prep_for_execute($op, $bind, $ident, \@args);
1524 $prepared_bind ||= [];
1527 ? ($sql, $prepared_bind, $bind_attrs)
1528 : \[ "($sql)", @$prepared_bind ]
1533 my ($self, $ident, $select, $where, $attrs) = @_;
1535 my ($alias2source, $rs_alias) = $self->_resolve_ident_sources ($ident);
1537 my $sql_maker = $self->sql_maker;
1538 $sql_maker->{_dbic_rs_attrs} = {
1544 ? ( _source_handle => $alias2source->{$rs_alias}->handle )
1549 # calculate bind_attrs before possible $ident mangling
1550 my $bind_attrs = {};
1551 for my $alias (keys %$alias2source) {
1552 my $bindtypes = $self->source_bind_attributes ($alias2source->{$alias}) || {};
1553 for my $col (keys %$bindtypes) {
1555 my $fqcn = join ('.', $alias, $col);
1556 $bind_attrs->{$fqcn} = $bindtypes->{$col} if $bindtypes->{$col};
1558 # Unqialified column names are nice, but at the same time can be
1559 # rather ambiguous. What we do here is basically go along with
1560 # the loop, adding an unqualified column slot to $bind_attrs,
1561 # alongside the fully qualified name. As soon as we encounter
1562 # another column by that name (which would imply another table)
1563 # we unset the unqualified slot and never add any info to it
1564 # to avoid erroneous type binding. If this happens the users
1565 # only choice will be to fully qualify his column name
1567 if (exists $bind_attrs->{$col}) {
1568 $bind_attrs->{$col} = {};
1571 $bind_attrs->{$col} = $bind_attrs->{$fqcn};
1578 $attrs->{software_limit}
1580 $sql_maker->_default_limit_syntax eq "GenericSubQ"
1582 $attrs->{software_limit} = 1;
1585 $self->throw_exception("rows attribute must be positive if present")
1586 if (defined($attrs->{rows}) && !($attrs->{rows} > 0));
1588 # MySQL actually recommends this approach. I cringe.
1589 $attrs->{rows} = 2**48 if not defined $attrs->{rows} and defined $attrs->{offset};
1594 # see if we need to tear the prefetch apart (either limited has_many or grouped prefetch)
1595 # otherwise delegate the limiting to the storage, unless software limit was requested
1597 ( $attrs->{rows} && keys %{$attrs->{collapse}} )
1599 ( $attrs->{group_by} && @{$attrs->{group_by}} &&
1600 $attrs->{_prefetch_select} && @{$attrs->{_prefetch_select}} )
1602 ($ident, $select, $where, $attrs)
1603 = $self->_adjust_select_args_for_complex_prefetch ($ident, $select, $where, $attrs);
1605 elsif (! $attrs->{software_limit} ) {
1606 push @limit, $attrs->{rows}, $attrs->{offset};
1610 # This would be the point to deflate anything found in $where
1611 # (and leave $attrs->{bind} intact). Problem is - inflators historically
1612 # expect a row object. And all we have is a resultsource (it is trivial
1613 # to extract deflator coderefs via $alias2source above).
1615 # I don't see a way forward other than changing the way deflators are
1616 # invoked, and that's just bad...
1620 { $attrs->{$_} ? ( $_ => $attrs->{$_} ) : () }
1621 (qw/order_by group_by having/ )
1624 return ('select', $attrs->{bind}, $ident, $bind_attrs, $select, $where, $order, @limit);
1628 # This is the code producing joined subqueries like:
1629 # SELECT me.*, other.* FROM ( SELECT me.* FROM ... ) JOIN other ON ...
1631 sub _adjust_select_args_for_complex_prefetch {
1632 my ($self, $from, $select, $where, $attrs) = @_;
1634 $self->throw_exception ('Nothing to prefetch... how did we get here?!')
1635 if not @{$attrs->{_prefetch_select}};
1637 $self->throw_exception ('Complex prefetches are not supported on resultsets with a custom from attribute')
1638 if (ref $from ne 'ARRAY' || ref $from->[0] ne 'HASH' || ref $from->[1] ne 'ARRAY');
1641 # generate inner/outer attribute lists, remove stuff that doesn't apply
1642 my $outer_attrs = { %$attrs };
1643 delete $outer_attrs->{$_} for qw/where bind rows offset group_by having/;
1645 my $inner_attrs = { %$attrs };
1646 delete $inner_attrs->{$_} for qw/for collapse _prefetch_select _collapse_order_by select as/;
1649 # bring over all non-collapse-induced order_by into the inner query (if any)
1650 # the outer one will have to keep them all
1651 delete $inner_attrs->{order_by};
1652 if (my $ord_cnt = @{$outer_attrs->{order_by}} - @{$outer_attrs->{_collapse_order_by}} ) {
1653 $inner_attrs->{order_by} = [
1654 @{$outer_attrs->{order_by}}[ 0 .. $ord_cnt - 1]
1659 # generate the inner/outer select lists
1660 # for inside we consider only stuff *not* brought in by the prefetch
1661 # on the outside we substitute any function for its alias
1662 my $outer_select = [ @$select ];
1663 my $inner_select = [];
1664 for my $i (0 .. ( @$outer_select - @{$outer_attrs->{_prefetch_select}} - 1) ) {
1665 my $sel = $outer_select->[$i];
1667 if (ref $sel eq 'HASH' ) {
1668 $sel->{-as} ||= $attrs->{as}[$i];
1669 $outer_select->[$i] = join ('.', $attrs->{alias}, ($sel->{-as} || "inner_column_$i") );
1672 push @$inner_select, $sel;
1675 # normalize a copy of $from, so it will be easier to work with further
1676 # down (i.e. promote the initial hashref to an AoH)
1678 $from->[0] = [ $from->[0] ];
1679 my %original_join_info = map { $_->[0]{-alias} => $_->[0] } (@$from);
1682 # decide which parts of the join will remain in either part of
1683 # the outer/inner query
1685 # First we compose a list of which aliases are used in restrictions
1686 # (i.e. conditions/order/grouping/etc). Since we do not have
1687 # introspectable SQLA, we fall back to ugly scanning of raw SQL for
1688 # WHERE, and for pieces of ORDER BY in order to determine which aliases
1689 # need to appear in the resulting sql.
1690 # It may not be very efficient, but it's a reasonable stop-gap
1691 # Also unqualified column names will not be considered, but more often
1692 # than not this is actually ok
1694 # In the same loop we enumerate part of the selection aliases, as
1695 # it requires the same sqla hack for the time being
1696 my ($restrict_aliases, $select_aliases, $prefetch_aliases);
1698 # produce stuff unquoted, so it can be scanned
1699 my $sql_maker = $self->sql_maker;
1700 local $sql_maker->{quote_char};
1701 my $sep = $self->_sql_maker_opts->{name_sep} || '.';
1704 my $non_prefetch_select_sql = $sql_maker->_recurse_fields ($inner_select);
1705 my $prefetch_select_sql = $sql_maker->_recurse_fields ($outer_attrs->{_prefetch_select});
1706 my $where_sql = $sql_maker->where ($where);
1707 my $group_by_sql = $sql_maker->_order_by({
1708 map { $_ => $inner_attrs->{$_} } qw/group_by having/
1710 my @non_prefetch_order_by_chunks = (map
1711 { ref $_ ? $_->[0] : $_ }
1712 $sql_maker->_order_by_chunks ($inner_attrs->{order_by})
1716 for my $alias (keys %original_join_info) {
1717 my $seen_re = qr/\b $alias $sep/x;
1719 for my $piece ($where_sql, $group_by_sql, @non_prefetch_order_by_chunks ) {
1720 if ($piece =~ $seen_re) {
1721 $restrict_aliases->{$alias} = 1;
1725 if ($non_prefetch_select_sql =~ $seen_re) {
1726 $select_aliases->{$alias} = 1;
1729 if ($prefetch_select_sql =~ $seen_re) {
1730 $prefetch_aliases->{$alias} = 1;
1736 # Add any non-left joins to the restriction list (such joins are indeed restrictions)
1737 for my $j (values %original_join_info) {
1738 my $alias = $j->{-alias} or next;
1739 $restrict_aliases->{$alias} = 1 if (
1740 (not $j->{-join_type})
1742 ($j->{-join_type} !~ /^left (?: \s+ outer)? $/xi)
1746 # mark all join parents as mentioned
1747 # (e.g. join => { cds => 'tracks' } - tracks will need to bring cds too )
1748 for my $collection ($restrict_aliases, $select_aliases) {
1749 for my $alias (keys %$collection) {
1750 $collection->{$_} = 1
1751 for (@{ $original_join_info{$alias}{-join_path} || [] });
1755 # construct the inner $from for the subquery
1756 my %inner_joins = (map { %{$_ || {}} } ($restrict_aliases, $select_aliases) );
1758 for my $j (@$from) {
1759 push @inner_from, $j if $inner_joins{$j->[0]{-alias}};
1762 # if a multi-type join was needed in the subquery ("multi" is indicated by
1763 # presence in {collapse}) - add a group_by to simulate the collapse in the subq
1764 unless ($inner_attrs->{group_by}) {
1765 for my $alias (keys %inner_joins) {
1767 # the dot comes from some weirdness in collapse
1768 # remove after the rewrite
1769 if ($attrs->{collapse}{".$alias"}) {
1770 $inner_attrs->{group_by} ||= $inner_select;
1776 # demote the inner_from head
1777 $inner_from[0] = $inner_from[0][0];
1779 # generate the subquery
1780 my $subq = $self->_select_args_to_query (
1787 my $subq_joinspec = {
1788 -alias => $attrs->{alias},
1789 -source_handle => $inner_from[0]{-source_handle},
1790 $attrs->{alias} => $subq,
1793 # Generate the outer from - this is relatively easy (really just replace
1794 # the join slot with the subquery), with a major caveat - we can not
1795 # join anything that is non-selecting (not part of the prefetch), but at
1796 # the same time is a multi-type relationship, as it will explode the result.
1798 # There are two possibilities here
1799 # - either the join is non-restricting, in which case we simply throw it away
1800 # - it is part of the restrictions, in which case we need to collapse the outer
1801 # result by tackling yet another group_by to the outside of the query
1803 # so first generate the outer_from, up to the substitution point
1805 while (my $j = shift @$from) {
1806 if ($j->[0]{-alias} eq $attrs->{alias}) { # time to swap
1811 last; # we'll take care of what's left in $from below
1814 push @outer_from, $j;
1818 # see what's left - throw away if not selecting/restricting
1819 # also throw in a group_by if restricting to guard against
1820 # cross-join explosions
1822 while (my $j = shift @$from) {
1823 my $alias = $j->[0]{-alias};
1825 if ($select_aliases->{$alias} || $prefetch_aliases->{$alias}) {
1826 push @outer_from, $j;
1828 elsif ($restrict_aliases->{$alias}) {
1829 push @outer_from, $j;
1831 # FIXME - this should be obviated by SQLA2, as I'll be able to
1832 # have restrict_inner and restrict_outer... or something to that
1833 # effect... I think...
1835 # FIXME2 - I can't find a clean way to determine if a particular join
1836 # is a multi - instead I am just treating everything as a potential
1837 # explosive join (ribasushi)
1839 # if (my $handle = $j->[0]{-source_handle}) {
1840 # my $rsrc = $handle->resolve;
1841 # ... need to bail out of the following if this is not a multi,
1842 # as it will be much easier on the db ...
1844 $outer_attrs->{group_by} ||= $outer_select;
1849 # demote the outer_from head
1850 $outer_from[0] = $outer_from[0][0];
1852 # This is totally horrific - the $where ends up in both the inner and outer query
1853 # Unfortunately not much can be done until SQLA2 introspection arrives, and even
1854 # then if where conditions apply to the *right* side of the prefetch, you may have
1855 # to both filter the inner select (e.g. to apply a limit) and then have to re-filter
1856 # the outer select to exclude joins you didin't want in the first place
1858 # OTOH it can be seen as a plus: <ash> (notes that this query would make a DBA cry ;)
1859 return (\@outer_from, $outer_select, $where, $outer_attrs);
1862 sub _resolve_ident_sources {
1863 my ($self, $ident) = @_;
1865 my $alias2source = {};
1868 # the reason this is so contrived is that $ident may be a {from}
1869 # structure, specifying multiple tables to join
1870 if ( Scalar::Util::blessed($ident) && $ident->isa("DBIx::Class::ResultSource") ) {
1871 # this is compat mode for insert/update/delete which do not deal with aliases
1872 $alias2source->{me} = $ident;
1875 elsif (ref $ident eq 'ARRAY') {
1879 if (ref $_ eq 'HASH') {
1881 $rs_alias = $tabinfo->{-alias};
1883 if (ref $_ eq 'ARRAY' and ref $_->[0] eq 'HASH') {
1887 $alias2source->{$tabinfo->{-alias}} = $tabinfo->{-source_handle}->resolve
1888 if ($tabinfo->{-source_handle});
1892 return ($alias2source, $rs_alias);
1895 # Takes $ident, \@column_names
1897 # returns { $column_name => \%column_info, ... }
1898 # also note: this adds -result_source => $rsrc to the column info
1901 # my $col_sources = $self->_resolve_column_info($ident, @column_names);
1902 sub _resolve_column_info {
1903 my ($self, $ident, $colnames) = @_;
1904 my ($alias2src, $root_alias) = $self->_resolve_ident_sources($ident);
1906 my $sep = $self->_sql_maker_opts->{name_sep} || '.';
1909 my (%return, %seen_cols);
1911 # compile a global list of column names, to be able to properly
1912 # disambiguate unqualified column names (if at all possible)
1913 for my $alias (keys %$alias2src) {
1914 my $rsrc = $alias2src->{$alias};
1915 for my $colname ($rsrc->columns) {
1916 push @{$seen_cols{$colname}}, $alias;
1921 foreach my $col (@$colnames) {
1922 my ($alias, $colname) = $col =~ m/^ (?: ([^$sep]+) $sep)? (.+) $/x;
1925 # see if the column was seen exactly once (so we know which rsrc it came from)
1926 if ($seen_cols{$colname} and @{$seen_cols{$colname}} == 1) {
1927 $alias = $seen_cols{$colname}[0];
1934 my $rsrc = $alias2src->{$alias};
1935 $return{$col} = $rsrc && {
1936 %{$rsrc->column_info($colname)},
1937 -result_source => $rsrc,
1938 -source_alias => $alias,
1945 # Returns a counting SELECT for a simple count
1946 # query. Abstracted so that a storage could override
1947 # this to { count => 'firstcol' } or whatever makes
1948 # sense as a performance optimization
1950 #my ($self, $source, $rs_attrs) = @_;
1951 return { count => '*' };
1954 # Returns a SELECT which will end up in the subselect
1955 # There may or may not be a group_by, as the subquery
1956 # might have been called to accomodate a limit
1958 # Most databases would be happy with whatever ends up
1959 # here, but some choke in various ways.
1961 sub _subq_count_select {
1962 my ($self, $source, $rs_attrs) = @_;
1963 return $rs_attrs->{group_by} if $rs_attrs->{group_by};
1965 my @pcols = map { join '.', $rs_attrs->{alias}, $_ } ($source->primary_columns);
1966 return @pcols ? \@pcols : [ 1 ];
1970 # Returns an ordered list of column names before they are used
1971 # in a SELECT statement. By default simply returns the list
1974 # This may be overridden in a specific storage when there are
1975 # requirements such as moving BLOB columns to the end of the
1977 sub _order_select_columns {
1978 #my ($self, $source, $columns) = @_;
1982 sub source_bind_attributes {
1983 my ($self, $source) = @_;
1985 my $bind_attributes;
1986 foreach my $column ($source->columns) {
1988 my $data_type = $source->column_info($column)->{data_type} || '';
1989 $bind_attributes->{$column} = $self->bind_attribute_by_data_type($data_type)
1993 return $bind_attributes;
2000 =item Arguments: $ident, $select, $condition, $attrs
2004 Handle a SQL select statement.
2010 my ($ident, $select, $condition, $attrs) = @_;
2011 return $self->cursor_class->new($self, \@_, $attrs);
2016 my ($rv, $sth, @bind) = $self->_select(@_);
2017 my @row = $sth->fetchrow_array;
2018 my @nextrow = $sth->fetchrow_array if @row;
2019 if(@row && @nextrow) {
2020 carp "Query returned more than one row. SQL that returns multiple rows is DEPRECATED for ->find and ->single";
2022 # Need to call finish() to work round broken DBDs
2031 =item Arguments: $sql
2035 Returns a L<DBI> sth (statement handle) for the supplied SQL.
2040 my ($self, $dbh, $sql) = @_;
2042 # 3 is the if_active parameter which avoids active sth re-use
2043 my $sth = $self->disable_sth_caching
2044 ? $dbh->prepare($sql)
2045 : $dbh->prepare_cached($sql, {}, 3);
2047 # XXX You would think RaiseError would make this impossible,
2048 # but apparently that's not true :(
2049 $self->throw_exception($dbh->errstr) if !$sth;
2055 my ($self, $sql) = @_;
2056 $self->dbh_do('_dbh_sth', $sql); # retry over disconnects
2059 sub _dbh_columns_info_for {
2060 my ($self, $dbh, $table) = @_;
2062 if ($dbh->can('column_info')) {
2065 my ($schema,$tab) = $table =~ /^(.+?)\.(.+)$/ ? ($1,$2) : (undef,$table);
2066 my $sth = $dbh->column_info( undef,$schema, $tab, '%' );
2068 while ( my $info = $sth->fetchrow_hashref() ){
2070 $column_info{data_type} = $info->{TYPE_NAME};
2071 $column_info{size} = $info->{COLUMN_SIZE};
2072 $column_info{is_nullable} = $info->{NULLABLE} ? 1 : 0;
2073 $column_info{default_value} = $info->{COLUMN_DEF};
2074 my $col_name = $info->{COLUMN_NAME};
2075 $col_name =~ s/^\"(.*)\"$/$1/;
2077 $result{$col_name} = \%column_info;
2080 return \%result if !$@ && scalar keys %result;
2084 my $sth = $dbh->prepare($self->sql_maker->select($table, undef, \'1 = 0'));
2086 my @columns = @{$sth->{NAME_lc}};
2087 for my $i ( 0 .. $#columns ){
2089 $column_info{data_type} = $sth->{TYPE}->[$i];
2090 $column_info{size} = $sth->{PRECISION}->[$i];
2091 $column_info{is_nullable} = $sth->{NULLABLE}->[$i] ? 1 : 0;
2093 if ($column_info{data_type} =~ m/^(.*?)\((.*?)\)$/) {
2094 $column_info{data_type} = $1;
2095 $column_info{size} = $2;
2098 $result{$columns[$i]} = \%column_info;
2102 foreach my $col (keys %result) {
2103 my $colinfo = $result{$col};
2104 my $type_num = $colinfo->{data_type};
2106 if(defined $type_num && $dbh->can('type_info')) {
2107 my $type_info = $dbh->type_info($type_num);
2108 $type_name = $type_info->{TYPE_NAME} if $type_info;
2109 $colinfo->{data_type} = $type_name if $type_name;
2116 sub columns_info_for {
2117 my ($self, $table) = @_;
2118 $self->_dbh_columns_info_for ($self->_get_dbh, $table);
2121 =head2 last_insert_id
2123 Return the row id of the last insert.
2127 sub _dbh_last_insert_id {
2128 # All Storage's need to register their own _dbh_last_insert_id
2129 # the old SQLite-based method was highly inappropriate
2132 my $class = ref $self;
2133 $self->throw_exception (<<EOE);
2135 No _dbh_last_insert_id() method found in $class.
2136 Since the method of obtaining the autoincrement id of the last insert
2137 operation varies greatly between different databases, this method must be
2138 individually implemented for every storage class.
2142 sub last_insert_id {
2144 $self->_dbh_last_insert_id ($self->_dbh, @_);
2147 =head2 _native_data_type
2151 =item Arguments: $type_name
2155 This API is B<EXPERIMENTAL>, will almost definitely change in the future, and
2156 currently only used by L<::AutoCast|DBIx::Class::Storage::DBI::AutoCast> and
2157 L<::Sybase|DBIx::Class::Storage::DBI::Sybase>.
2159 The default implementation returns C<undef>, implement in your Storage driver if
2160 you need this functionality.
2162 Should map types from other databases to the native RDBMS type, for example
2163 C<VARCHAR2> to C<VARCHAR>.
2165 Types with modifiers should map to the underlying data type. For example,
2166 C<INTEGER AUTO_INCREMENT> should become C<INTEGER>.
2168 Composite types should map to the container type, for example
2169 C<ENUM(foo,bar,baz)> becomes C<ENUM>.
2173 sub _native_data_type {
2174 #my ($self, $data_type) = @_;
2178 # Check if placeholders are supported at all
2179 sub _placeholders_supported {
2181 my $dbh = $self->_get_dbh;
2183 # some drivers provide a $dbh attribute (e.g. Sybase and $dbh->{syb_dynamic_supported})
2184 # but it is inaccurate more often than not
2186 local $dbh->{PrintError} = 0;
2187 local $dbh->{RaiseError} = 1;
2188 $dbh->do('select ?', {}, 1);
2193 # Check if placeholders bound to non-string types throw exceptions
2195 sub _typeless_placeholders_supported {
2197 my $dbh = $self->_get_dbh;
2200 local $dbh->{PrintError} = 0;
2201 local $dbh->{RaiseError} = 1;
2202 # this specifically tests a bind that is NOT a string
2203 $dbh->do('select 1 where 1 = ?', {}, 1);
2210 Returns the database driver name.
2217 if (not $self->_driver_determined) {
2218 $self->_determine_driver;
2219 goto $self->can ('sqlt_type');
2222 $self->_get_dbh->{Driver}->{Name};
2225 =head2 bind_attribute_by_data_type
2227 Given a datatype from column info, returns a database specific bind
2228 attribute for C<< $dbh->bind_param($val,$attribute) >> or nothing if we will
2229 let the database planner just handle it.
2231 Generally only needed for special case column types, like bytea in postgres.
2235 sub bind_attribute_by_data_type {
2239 =head2 is_datatype_numeric
2241 Given a datatype from column_info, returns a boolean value indicating if
2242 the current RDBMS considers it a numeric value. This controls how
2243 L<DBIx::Class::Row/set_column> decides whether to mark the column as
2244 dirty - when the datatype is deemed numeric a C<< != >> comparison will
2245 be performed instead of the usual C<eq>.
2249 sub is_datatype_numeric {
2250 my ($self, $dt) = @_;
2252 return 0 unless $dt;
2254 return $dt =~ /^ (?:
2255 numeric | int(?:eger)? | (?:tiny|small|medium|big)int | dec(?:imal)? | real | float | double (?: \s+ precision)? | (?:big)?serial
2260 =head2 create_ddl_dir (EXPERIMENTAL)
2264 =item Arguments: $schema \@databases, $version, $directory, $preversion, \%sqlt_args
2268 Creates a SQL file based on the Schema, for each of the specified
2269 database engines in C<\@databases> in the given directory.
2270 (note: specify L<SQL::Translator> names, not L<DBI> driver names).
2272 Given a previous version number, this will also create a file containing
2273 the ALTER TABLE statements to transform the previous schema into the
2274 current one. Note that these statements may contain C<DROP TABLE> or
2275 C<DROP COLUMN> statements that can potentially destroy data.
2277 The file names are created using the C<ddl_filename> method below, please
2278 override this method in your schema if you would like a different file
2279 name format. For the ALTER file, the same format is used, replacing
2280 $version in the name with "$preversion-$version".
2282 See L<SQL::Translator/METHODS> for a list of values for C<\%sqlt_args>.
2283 The most common value for this would be C<< { add_drop_table => 1 } >>
2284 to have the SQL produced include a C<DROP TABLE> statement for each table
2285 created. For quoting purposes supply C<quote_table_names> and
2286 C<quote_field_names>.
2288 If no arguments are passed, then the following default values are assumed:
2292 =item databases - ['MySQL', 'SQLite', 'PostgreSQL']
2294 =item version - $schema->schema_version
2296 =item directory - './'
2298 =item preversion - <none>
2302 By default, C<\%sqlt_args> will have
2304 { add_drop_table => 1, ignore_constraint_names => 1, ignore_index_names => 1 }
2306 merged with the hash passed in. To disable any of those features, pass in a
2307 hashref like the following
2309 { ignore_constraint_names => 0, # ... other options }
2312 Note that this feature is currently EXPERIMENTAL and may not work correctly
2313 across all databases, or fully handle complex relationships.
2315 WARNING: Please check all SQL files created, before applying them.
2319 sub create_ddl_dir {
2320 my ($self, $schema, $databases, $version, $dir, $preversion, $sqltargs) = @_;
2322 if(!$dir || !-d $dir) {
2323 carp "No directory given, using ./\n";
2326 $databases ||= ['MySQL', 'SQLite', 'PostgreSQL'];
2327 $databases = [ $databases ] if(ref($databases) ne 'ARRAY');
2329 my $schema_version = $schema->schema_version || '1.x';
2330 $version ||= $schema_version;
2333 add_drop_table => 1,
2334 ignore_constraint_names => 1,
2335 ignore_index_names => 1,
2339 $self->throw_exception(q{Can't create a ddl file without SQL::Translator 0.09003: '}
2340 . $self->_check_sqlt_message . q{'})
2341 if !$self->_check_sqlt_version;
2343 my $sqlt = SQL::Translator->new( $sqltargs );
2345 $sqlt->parser('SQL::Translator::Parser::DBIx::Class');
2346 my $sqlt_schema = $sqlt->translate({ data => $schema })
2347 or $self->throw_exception ($sqlt->error);
2349 foreach my $db (@$databases) {
2351 $sqlt->{schema} = $sqlt_schema;
2352 $sqlt->producer($db);
2355 my $filename = $schema->ddl_filename($db, $version, $dir);
2356 if (-e $filename && ($version eq $schema_version )) {
2357 # if we are dumping the current version, overwrite the DDL
2358 carp "Overwriting existing DDL file - $filename";
2362 my $output = $sqlt->translate;
2364 carp("Failed to translate to $db, skipping. (" . $sqlt->error . ")");
2367 if(!open($file, ">$filename")) {
2368 $self->throw_exception("Can't open $filename for writing ($!)");
2371 print $file $output;
2374 next unless ($preversion);
2376 require SQL::Translator::Diff;
2378 my $prefilename = $schema->ddl_filename($db, $preversion, $dir);
2379 if(!-e $prefilename) {
2380 carp("No previous schema file found ($prefilename)");
2384 my $difffile = $schema->ddl_filename($db, $version, $dir, $preversion);
2386 carp("Overwriting existing diff file - $difffile");
2392 my $t = SQL::Translator->new($sqltargs);
2397 or $self->throw_exception ($t->error);
2399 my $out = $t->translate( $prefilename )
2400 or $self->throw_exception ($t->error);
2402 $source_schema = $t->schema;
2404 $source_schema->name( $prefilename )
2405 unless ( $source_schema->name );
2408 # The "new" style of producers have sane normalization and can support
2409 # diffing a SQL file against a DBIC->SQLT schema. Old style ones don't
2410 # And we have to diff parsed SQL against parsed SQL.
2411 my $dest_schema = $sqlt_schema;
2413 unless ( "SQL::Translator::Producer::$db"->can('preprocess_schema') ) {
2414 my $t = SQL::Translator->new($sqltargs);
2419 or $self->throw_exception ($t->error);
2421 my $out = $t->translate( $filename )
2422 or $self->throw_exception ($t->error);
2424 $dest_schema = $t->schema;
2426 $dest_schema->name( $filename )
2427 unless $dest_schema->name;
2430 my $diff = SQL::Translator::Diff::schema_diff($source_schema, $db,
2434 if(!open $file, ">$difffile") {
2435 $self->throw_exception("Can't write to $difffile ($!)");
2443 =head2 deployment_statements
2447 =item Arguments: $schema, $type, $version, $directory, $sqlt_args
2451 Returns the statements used by L</deploy> and L<DBIx::Class::Schema/deploy>.
2453 The L<SQL::Translator> (not L<DBI>) database driver name can be explicitly
2454 provided in C<$type>, otherwise the result of L</sqlt_type> is used as default.
2456 C<$directory> is used to return statements from files in a previously created
2457 L</create_ddl_dir> directory and is optional. The filenames are constructed
2458 from L<DBIx::Class::Schema/ddl_filename>, the schema name and the C<$version>.
2460 If no C<$directory> is specified then the statements are constructed on the
2461 fly using L<SQL::Translator> and C<$version> is ignored.
2463 See L<SQL::Translator/METHODS> for a list of values for C<$sqlt_args>.
2467 sub deployment_statements {
2468 my ($self, $schema, $type, $version, $dir, $sqltargs) = @_;
2469 $type ||= $self->sqlt_type;
2470 $version ||= $schema->schema_version || '1.x';
2472 my $filename = $schema->ddl_filename($type, $version, $dir);
2476 open($file, "<$filename")
2477 or $self->throw_exception("Can't open $filename ($!)");
2480 return join('', @rows);
2483 $self->throw_exception(q{Can't deploy without SQL::Translator 0.09003: '}
2484 . $self->_check_sqlt_message . q{'})
2485 if !$self->_check_sqlt_version;
2487 # sources needs to be a parser arg, but for simplicty allow at top level
2489 $sqltargs->{parser_args}{sources} = delete $sqltargs->{sources}
2490 if exists $sqltargs->{sources};
2492 my $tr = SQL::Translator->new(
2493 producer => "SQL::Translator::Producer::${type}",
2495 parser => 'SQL::Translator::Parser::DBIx::Class',
2498 return $tr->translate;
2502 my ($self, $schema, $type, $sqltargs, $dir) = @_;
2505 return if($line =~ /^--/);
2507 # next if($line =~ /^DROP/m);
2508 return if($line =~ /^BEGIN TRANSACTION/m);
2509 return if($line =~ /^COMMIT/m);
2510 return if $line =~ /^\s+$/; # skip whitespace only
2511 $self->_query_start($line);
2513 # do a dbh_do cycle here, as we need some error checking in
2514 # place (even though we will ignore errors)
2515 $self->dbh_do (sub { $_[1]->do($line) });
2518 carp qq{$@ (running "${line}")};
2520 $self->_query_end($line);
2522 my @statements = $self->deployment_statements($schema, $type, undef, $dir, { %{ $sqltargs || {} }, no_comments => 1 } );
2523 if (@statements > 1) {
2524 foreach my $statement (@statements) {
2525 $deploy->( $statement );
2528 elsif (@statements == 1) {
2529 foreach my $line ( split(";\n", $statements[0])) {
2535 =head2 datetime_parser
2537 Returns the datetime parser class
2541 sub datetime_parser {
2543 return $self->{datetime_parser} ||= do {
2544 $self->_populate_dbh unless $self->_dbh;
2545 $self->build_datetime_parser(@_);
2549 =head2 datetime_parser_type
2551 Defines (returns) the datetime parser class - currently hardwired to
2552 L<DateTime::Format::MySQL>
2556 sub datetime_parser_type { "DateTime::Format::MySQL"; }
2558 =head2 build_datetime_parser
2560 See L</datetime_parser>
2564 sub build_datetime_parser {
2566 my $type = $self->datetime_parser_type(@_);
2568 $self->throw_exception("Couldn't load ${type}: $@") if $@;
2573 my $_check_sqlt_version; # private
2574 my $_check_sqlt_message; # private
2575 sub _check_sqlt_version {
2576 return $_check_sqlt_version if defined $_check_sqlt_version;
2577 eval 'use SQL::Translator "0.09003"';
2578 $_check_sqlt_message = $@ || '';
2579 $_check_sqlt_version = !$@;
2582 sub _check_sqlt_message {
2583 _check_sqlt_version if !defined $_check_sqlt_message;
2584 $_check_sqlt_message;
2588 =head2 is_replicating
2590 A boolean that reports if a particular L<DBIx::Class::Storage::DBI> is set to
2591 replicate from a master database. Default is undef, which is the result
2592 returned by databases that don't support replication.
2596 sub is_replicating {
2601 =head2 lag_behind_master
2603 Returns a number that represents a certain amount of lag behind a master db
2604 when a given storage is replicating. The number is database dependent, but
2605 starts at zero and increases with the amount of lag. Default in undef
2609 sub lag_behind_master {
2615 $self->_verify_pid if $self->_dbh;
2617 # some databases need this to stop spewing warnings
2618 if (my $dbh = $self->_dbh) {
2619 eval { $dbh->disconnect };
2629 =head2 DBIx::Class and AutoCommit
2631 DBIx::Class can do some wonderful magic with handling exceptions,
2632 disconnections, and transactions when you use C<< AutoCommit => 1 >>
2633 (the default) combined with C<txn_do> for transaction support.
2635 If you set C<< AutoCommit => 0 >> in your connect info, then you are always
2636 in an assumed transaction between commits, and you're telling us you'd
2637 like to manage that manually. A lot of the magic protections offered by
2638 this module will go away. We can't protect you from exceptions due to database
2639 disconnects because we don't know anything about how to restart your
2640 transactions. You're on your own for handling all sorts of exceptional
2641 cases if you choose the C<< AutoCommit => 0 >> path, just as you would
2647 Matt S. Trout <mst@shadowcatsystems.co.uk>
2649 Andy Grundman <andy@hybridized.org>
2653 You may distribute this code under the same terms as Perl itself.